The goals of this proposal are to identify distinctive phenotypic characteristics of malignant human astrocytes and to utilize these biochemical markers for diagnostic classification of astrocytomas. The working hypothesis of the project is that differentiation-specific markers for early stages of normal astrocyte development are re-expressed in malignant astrocytes and can therefore serve to identify malignant phenotypes. Preliminary studies have shown that proteins of the intermediate filament (IF) cytoskeleton occupy unique temporal niches in astrocyte development. One such differentiation marker, IF-associated protein (IFAP)-3OOkD, appears to be common to all human astrocytomas, while keratin IF protein production appears to typify the most malignant grade of astrocytomas. In this context, the Specific Aims of this revised proposal are: 1) to define the expression of IFAP-3OOkD as a putative marker of malignant human astrocytes; 2) to biochemically characterize IFAP-3OOkD from human astrocytoma cells; 3) to produce alternative antibodies to IFAP-3OOkD in order to study its expression using fixed human astrocytoma specimens; 4) to investigate keratin expression in human astrocytomas as an index of tumor malignancy; 5) to survey the spectrum of tumor types with a library of existing monoclonal antibodies to rat astrocyte IF-related proteins in order to identify other markers of the malignant phenotype; 6) to produce monoclonal antibodies against IF-related proteins obtained directly from human astrocytomas. The basis of the expression patterns for the various protein markers will be determined directly in human astrocytoma tissue specimens by immunofluorescence (and immunoperoxidase) microscopy using previously developed monoclonal antibodies. Specimens from normal human CNS (for control) as well as from human CNS degenerative diseases (for reactive astrocytes) will be included. The polypeptide composition of the markers will be determined by SDS-PAGE and 2-D PAGE/inmmunoblot analysis. IFAP-3OOkD will be chromatographically purified from astrocytomas and characterized by amino acid analysis, peptide mapping, and N-terminal sequence analysis. The purified IFAP-3OOkD will also be denatured and used to produce new antibodies so that aldehyde-fixed specimens can be surveyed. Partial biochemical characterization of the keratin cross-reactive species may be necessary as a means of identification relative to the known keratin polypeptides. The tissue specimens will be additionally surveyed with the library of other monoclonal antibodies by immunofluorescence microscopy and PAGE/immunoblot analysis. Finally, after PAGE comparison of normal and malignant astrocytic cell IF protein profiles, proteins unique to the malignant phenotype will be purified for the production of monoclonal antibodies which will then be screened to identify other markers of astrocytomas.